Catalytic-doping of silicon alloys for the use in silicon heterojunction solar cells

To make this planet cleaner and more sustainable, photovoltaic (PV) has long been on its road to pursue higher efficiency and larger market share. Silicon heterojunction (SHJ) solar cell is considered as one of the most promising candidates as next-generation high-efficiency PV techniques due to their high efficiencies and simple production sequences. Efficiency over 26% in research level and over 23% in mass production have been achieved based on SHJ concept. However, there are still several efficiency-limiting issues for SHJ solar cells. The first one is the parasitic absorption in short wavelength range caused by front side doped amorphous (a-Si:H)layers. Many new materials (such as uc-Si:H, nc-SiOx:H and uc-SiC:H) have been proposed to overcome this problem, yet the trade-off between optical and electrical properties still exists. The second one is that sufficient thickness of a-Si:H(i) layer is needed for good passivation, but the a-Si:H(i) layer acts as isolation for electrical transport. Therefore, another trade-off exists between passivation and electrical transport. Among many techniques to solve aforementioned trade-offs and even push efficiency forward, Catalytic-doping (Cat-doping) is a cheap, simple and industry feasible technique. Cat-doping has been reported to improve electrical property and passivation quality on films and device consisting of a-Si:H and crystalline silicon(c-Si). However, the effectiveness of Cat-doping on silicon alloys ( uc-Si:H, nc-SiOx:H and uc-SiC:H ) is still unknown. The mechanism of Cat-doping on these materials is also not clear so far. Accordingly, the first goal of this thesis was to exploit the effectiveness of Cat-doping on these silicon alloy materials in terms of electrical and optical properties. Meanwhile, deeper understanding of the mechanism of Cat-doping in different materials was needed. It is shown that P atoms can be effectively incorporated into these films by Cat-doping but with different P atom concentrations (NP) and doping depths. As for a-Si:H films, NP from Cat-doping can be as high as 6x10^21cm^3 with a doping depth around 15 nm while for uc-Si:H NP can reach 6x10^20cm^3 with a doping depth of 5 nm. For nc-SiOx:H films, NP is also in the range of 10^20cm^3 but P atoms penetrate throughout the whole film. When it comes to uc-SiC:H films, NP is only 6x10^19cm^3 and P atoms seem to just attach to the surface of uc-SiC:H films. The difference in the behavior of Cat-doping might be attributed to the microstructures of these silicon alloys. The loose structure of a-Si:H, where a large number of voids and defects exist, possibly results in large doping concentration NP and deeper doping depth for a-Si:H. The grain boundaries in uc-Si:H could also help to incorporate more P atoms. Nanocrystalline silicon oxide has a very unique porous cage microstructure that might offer channels for P atoms to penetrate the whole film during the Cat-doping process. In contrast, the dense microstructure of uc-SiC:H seems to prevent P atoms to penetrate and the P atoms only attach to the film surface. The second goal of this work was to improve the passivation quality of different cell precursors and devices. To begin with, different gases were used during Cat-doping process and three effects (i.e. thermal annealing, hydrogenation and P doping effect) were found to co-exist accounting for the improvement of passivation quality. Then, cell precursors and devices with these silicon alloys films were treated with Cat-doping and their passivation quality were enhanced to different extend, which is also likely related to the change in material properties upon applying the Cat-doping process. At last, the ultimate goal was to implement Cat-doping in SHJ solar cells. In general, applying Cat-doping on various doped layers in the front of SHJ solar cells results in improved efficiency which mainly comes from improved Voc and FF. Filament temperature (Tf), substrate temperature (Tsub) and processing pressure were all found to influence efficiency and local optimal values were identified. However, the electrical property change for nc-SiOx:H seems limited due to the low activation of Cat-doping incorporated P atoms, the impact of Cat-doping on efficiency of cells with nc-SiOx:H was not significant. Beyond the traditional structure of SHJ cells, a new fabrication method where the deposition of a-Si:H(n) layer was replaced by Cat-doping on a-Si:H(i) layer was proposed. Compared to traditional SHJ cell, equivalent efficiency from cells with new structure was achieved and has demonstrated an increase in Jsc about ~ 1mA/cm^2. Based on this concept, when additional a-Si:H(n) layer was added to the front of the new-approach cell, a champion device of 21.2% efficiency was achieved mainly due to an increase of 4.9% in FF compared to the traditional SHJ cell structure